This dissertation aimed to develop and validate a need-supportive and -thwarting instruction observation scale and perform psychometric testing of the validity and reliability of the scale to assess instructors’ instructional behaviors in the physical activity context. An additional aim of the dissertation was to compare the effects of SDT-centered need-supportive and -thwarting instruction on participants’ motivational, behavioral, and affective outcomes. Developing a need-supportive and -thwarting instruction observation scale study showed that between-rater correlation (.80 to .95) suggested strong internal consistency. Factorial validity was excellent, factor loading ranging between .88 and .99, with 92% (CI 95% [.89, .95]) of the total variance explained. Confirmatory factor analysis supported the one-factor solution (χ2(27) = 35.45; CFI = .99; TLI = .99; RMSEA = .07). The findings showed that the need-supportive and -thwarting instruction observation scale had the desired validity and reliability to measure instructor’s need-supportive and -thwarting behaviors in physical activity. The cluster-randomized trial showed that need-supportive instruction improved participants’ skill performance (F[2,106] = 4.15, p = .044, ηp2 = .04), intrinsic motivation (F[2,106] = 8.87, p = .004, ηp2 = .08), introjected regulation (F[2,106] = 6.06, p = .012, ηp2 = .06), extrinsic regulation (F[2,106] = 9.45, p = .003, ηp2= .08), and amotivation (F[2,106] = 12.29, p < .001, ηp2 = .12). Also, there was a significant between-group effect on enjoyment (F[2,106] = 24.23, p < .001, ηp2 = .19) and state anxiety (F[2,106] = 3.17, p = .020, ηp2= .05). The results suggest that the need-supportive instruction was a superior instructional strategy compared to need-thwarting instruction when aiming to improve participants’ motor skill performance, motivation, and positive affect.